Data Sheet 1_Establishment and validation of a CT-based clinical deep learning radiomics n

A novel clinical deep learning radiomics nomogram (CDLRN) achieved an area under the curve (AUC) of 0.958 in training and 0.916 in validation for predicting transcatheter arterial chemoembolization (TACE) response. The model was developed by Genxiang Chen using pre-treatment CT images and clinical data from 144 HCC patients treated between January 2019 and December 2023. Its specificity reached 1.0 and accuracy was 83.3% in the validation cohort.

Use Cases

• Predicting hepatocellular carcinoma (HCC) patient response to transcatarterial chemoembolization (TACE) based on clinical and CT radiomic features.
• Training and validating deep learning models for medical image analysis using contrast-enhanced CT scans.
• Building clinical decision support tools for personalized cancer treatment planning based on the described nomogram methodology.

Strengths

• Model validation showed high performance with an AUC of 0.916 in a separate cohort of 42 patients.
• The dataset includes clinical data and pre-treatment CT images from 144 patients, providing a multimodal foundation.
• Response was assessed using the standardized Modified Response Evaluation Criteria in Solid Tumors (mRECIST).

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data is from a single hospital (Shangluo Central Hospital), which may reflect geographic or institutional bias.

Provenance

Source

figshare, author Genxiang Chen.

Collection Method

Clinical and CT data were retrospectively collected from HCC patients who underwent TACE.

Time Range

January 2019 to December 2023.

Freshness

Last updated 2026-04-22 11:01:31; freshness should be verified.

Geography

Patients treated at the Department of Hepatobiliary Surgery of Shangluo Central Hospital (location not fully specified).